Graphics system with an improved filtering adder tree

Abstract

A sample-to-pixel calculation unit in a graphics system may comprise an adder tree. The adder tree includes a plurality of adder cells coupled in a tree configuration. Input values are presented to a first layer of adder cells. Each input value may have two associated control signals: a data valid signal and a winner-take-all signal. The final output of the adder tree equals (a) a sum of those input values whose data valid signals are asserted provided that none of the winner-take-all signals are asserted, or (b) a selected one of the input values if one of the winner-take-all bits is asserted. The selected input value is the one whose winner-take-all bit is set. The adder tree may be used to perform sums of weighted sample attributes and / or sums of coefficients values as part of pixel value computations.

Description

I. CONTINUATION DATA[0001]This application claims benefit of priority to U.S. Provisional Application Ser. No. 60 / 215,030 filed on Jun. 29, 2000 titled “Graphics System with an Improved Filtering Adder Tree”.II. BACKGROUND OF THE INVENTION[0002]1. Field of the Invention[0003]This invention relates generally to the field of computer graphics and, more particularly, to high performance graphics systems.[0004]2. Description of the Related Art[0005]A host computer may rely on a graphics system for producing visual output on a display device. The graphics system may receive graphics data (e.g. triangle data) from the host computer, and may generate a stream of output pixels in response to the graphics data. The stream of output pixels may be stored in a frame buffer, and then dumped from the frame buffer to a display device such as a monitor or projection screen.[0006]To obtain images that are more realistic, some prior art graphics systems have gone further by generating more than one s...

AI Technical Summary

Benefits of technology

[0029]In cases where the filter support covers regions with two or more different sample densities, the samples from the lower density regions may contribute less to the final pixel value than the samples from the higher density region. This is because there are typically fewer samples in the lower density region. In one embodiment, the filter coefficients corresponding to samples from the lower sample density regions may be multiplied by a factor approximately equal to the ratio of the sample densities. This may provide more weight to the less-represented samples from the lower density region. In cases where the filter support may include more than two regions of different sample densities, filter coefficients for samples in other regions may also be multiplied by a factor equal to the ratio of the sample densities.

Problems solved by technology

These systems, however, have generally suffered from limitations imposed by the conventional frame buffer and by the added latency caused by the render buffer and filtering.

However, one potential obstacle to an improved graphics system is that the filtering operation may be computationally intensive.

This typically translates into a large number of calculations.

Images